Saturday, 14 April 2018

ADS-B Groundplane antenna - version 2

I finally had chance to analyze the return loss and bandwidth of the wire groundplane ADS-B antenna recently, and found that it was actually tuned to 1300MHz! As this is higher in frequency than the required 1090MHz, this meant that the antenna elements were too short. Had the frequency been low, then it would have been a simple case of pruning the elements a millimeter or so at a time to frequency, but being too short means they need extending, which is all but impossible due to the enamel insulation.

So I decided that I would build a new antenna, this time deliberately too low in frequency, to allow me to prune it. I also opted to use bare 1.5mm2 copper wire, which will allow me to add length if required.

One of the most difficult aspects of the original antenna was the mounting disk for the ground plane elements. On the original this is made from PCB stock, but I wanted something that was truly circular for this one. As luck would have it, a visit to the National Coal Mining Museum at Caphouse Colliery supplied me with the perfect starting material - a brass pit check!

Caphouse Colliery pit check

 Finding the center of the disk as somewhat tricky, and even then, the hole isnt perfectly centered, but it also is a little large, so I was able to adjust the BNC socket position on the disk to be more accurately centered later.

The wire im using here is 1.5mm2 domestic Twin and Earth mains cable, stripped of its insulation. The pre-stamped small hole in the pit check, intended for hanging it on the banksmans board in the lamp room of the pit head, was perfect for allowing a small woodscrew to hold the disk in place, while ordinary masking tape held the wire elements steady ready for soldering. The elements here are cut to 85mm, allowing 5mm for soldering.

Cardinal point elements in place

With the first four elements in place, the next four were soldered in. Being a solid brass disk, and with the aid of my new 150W soldering iron, the joints were made quickly and cleanly.

Remaining elements fitted

Once cool, the assembled ground plane was removed from the wooden board, and the vertical element added. This was cut initially to 100mm, then reduced down to 90mm once soldered and straightened.
First Test

 I have yet to analyze the properties of this build, and will adjust the ground plane element angles down to achieve the correct impedance at that time. For now, a quick test of it on the ADS-B receiver showed that it is working, and seems to be working as well as the original antenna.


On each blog post  I will be including the following link -
 £50 credit if you switch energy supplier to Bulb Ltd
This is my own personal referral link, not an advert! I am with Bulb for gas and electricity. Do your own research, but if you decide to switch to them, do so via the above link, and we both get a £50 credit!


Saturday, 24 February 2018

Coupler?

This post only exists to allow me a way of passing a particular photo.

Coupler -




On each blog post  I will be including the following link -
 £50 credit if you switch energy supplier to Bulb Ltd
This is my own personal referral link, not an advert! I am with Bulb for gas and electricity. Do your own research, but if you decide to switch to them, do so via the above link, and we both get a £50 credit!

Thursday, 22 February 2018

Splat!

After returning home from work this morning, following a hard session at the gym, the ADS-B aggregation had done a little over 24h, allowing time to receive squitters from a whole weekdays worth of scheduled, charter, military and general aviation flights, plus all the overnight cargo flights.

The resulting aggregated plot, known as a 'splat', for reasons which should become apparent, for the 1090MHz Ground Plane antenna, mounted on Sams south facing windowsill, is shown below


It is clear that coverage to the north in this situation is very sketchy, and to the west very poor due to having to also pass through the bulk of the neighbours property as well. The best coverage area, both for high altitude (red, purple) and low level (greens) is unsurprisingly to the south-west, this being the visually least obstructed view from Sams window.

I need to say many thanks to my old colleague Mike, who has supplied me with another Raspberry Pi SBC for Sam, to replace the one I am pushing into service for this project. Sams Pi is a model B+. I expected the one from Mike to be either a B+ or one of the older models, but to my great surprise he's sent a 3B! This is the current model, which includes WiFi, and is faster and has more RAM than the B+.  Im sure Sam will put it to good use. Mike also sent a Microbit SBC, a device I had never had my hands on, but Sam has used at school. And, if all that wasnt great enough - my Pi Zero W was delivered as well! Lots of great embedded processing options available to us now!

Whilst all this has been going on, ive also been building a 1:72 scale model of a Sopwith Camel! Ive just put the undercarriage on, which just leaves the upper wing for tomorrow!

In my previous post I showed a layout plan of the ADS-B system proposal. One thing that occurs to me is that since the system will be installed in a plastic tube, and the electronics mounted on a board slotted up the middle, there is no reason to try and fit all the electronics on the same side of the carrier! The carrier board effectively divides the tube into two semicircular spaces, so the Pi and RTL could go on one side, and the USB splitter, PoE splitter and Buck Converter on the other.



On each blog post  I will be including the following link -
 £50 credit if you switch energy supplier to Bulb Ltd
This is my own personal referral link, not an advert! I am with Bulb for gas and electricity. Do your own research, but if you decide to switch to them, do so via the above link, and we both get a £50 credit!

Mast-head ADS-B system - layout

More for my own purposes than anything else, just so I get my thoughts fully straight over what is required to be included in the systems radome, I have drawn up a diagram of the internal arrangement of the project

The antenna represented is the 1090MHz ground-plane. Although this has no dBd gain, it is however compact and effective when mounted clear of obstructions. The diagram only shows the necessary electrical connectivity, it does not in any way represent the hardware challenges involved.

At present, with the current test bed setup and its windowsill mounted antenna, I am running a data aggregation for 24h, to create what is termed a 'splat' - a graphical map based representation of the antennas coverage. I will post up the result of this once aggregation is complete.




On each blog post  I will be including the following link -
 £50 credit if you switch energy supplier to Bulb Ltd
This is my own personal referral link, not an advert! I am with Bulb for gas and electricity. Do your own research, but if you decide to switch to them, do so via the above link, and we both get a £50 credit!

Wednesday, 21 February 2018

The Raspberry Pi Range

No real reason for this post other than just to show the full range of Raspberry Pi computers. Im not gonna list the specs or anything. You can find them yourself online.


The original type A, the 2nd generation A+, and the B range


The 3, with WiFi etc


Not commonly seen by experimenters or educators, the computation unit is available without the peripherals



And the latest of the range, the Zero and Zero W


On each blog post  I will be including the following link -
 £50 credit if you switch energy supplier to Bulb Ltd
This is my own personal referral link, not an advert! I am with Bulb for gas and electricity. Do your own research, but if you decide to switch to them, do so via the above link, and we both get a £50 credit!

Antenna 'Master'!

Whilst looking at images for ADS-B antennas this evening on Google, I spotted my own groundplane design with the hairpin match, but it wasnt a link to this blog! It was this discussion here https://discussions.flightaware.com/t/built-my-first-antenna-and-doubled-my-coverage/15770/241 where my antenna was discussed with the mention that it was built 'by a master'!

Sadly not! However, it has prompted me to look at how to use my spectrum analyser and tracking generator to test these antennas, since my antenna analyser only goes up to 160MHz! Looks like I need to get a directional coupler!



On each blog post  I will be including the following link -
 £50 credit if you switch energy supplier to Bulb Ltd
This is my own personal referral link, not an advert! I am with Bulb for gas and electricity. Do your own research, but if you decide to switch to them, do so via the above link, and we both get a £50 credit!

Tuesday, 20 February 2018

Plans afoot for Mast-head ADS-B system

The Raspberry Pi 2B based ADS-B receiver system has now been operating succesfully for over 24h, feeding data into the 360radar servers, and so I am now an official contributor and have my log-in's for access to the system.

So the next task is to work to rebuild the receiver system into a self contained system that can be mounted in its entirety at the mast-head, in order to gain the clearest view of the sky, for the furthest range, and at the lowest possible signal loss.

There are a number of difficulties with this, but principly they come down to power feed and data feed.  WiFi could be used, and indeed I have ordered a Raspberry Pi ZeroW to play with this option. The Pi zeroW is a smaller board than the Pi 2/2B or 3, but with less physical connectivity, it only has one USB micro port, unlike the 2B with its four standard USB sockets, and it has no Ethernet port. But it does have on-board WiFi and Bluetooth.

The power feed is the other issue. If using WiFi, then just a DC supply is needed. But the cost of suitably rated DC cable is no cheaper than exterior grade CAT5, and Ethernet would be a more reliable data connection than WiFi (and more secure). This brings up the possibility of using PoE (Power over Ethernet) as a convenient means of supplying the system with power.

I did briefly toy with an alternative board, the Orange Pi, as this is smaller than the R-Pi and could apparently do PoE 'natively'. Sounded good, but the chipsets are different which would have meant much difficulty in compiling the code. This idea ive abandoned once I discovered that the 'native' PoE was simply LAN pins 4 and 5 being available at solder pads. No buck converter or low drop-out 5V regulator is provided, meaning that to provide any meaningful PoE capability I would have to add a Buck Converter anyway! So its easier to stay with the R-Pi and use PoE injectors/splitters.

A colleague has sent me another R-Pi, not sure which model as yet, in order that I can release Sams 2B back to him, rather than embed it in the system. Many thanks Mike! Im looking forward to receiving this unit and seeing what modifications are needed, if any, to my plan.

So, on then to the plan! The plan is cunning... and simplistic. The R-Pi, RTL dongle, Buck Converter, and Antenna, will all be enclosed within a suitable weather radome. Lets look at the total system, from air-side to data link side -

The antenna, which is built on the body of a chassis mount BNC socket, is connected via a BNC patch lead to a BNC to MCX pigtail cable. To reduce losses in future I may replace the patch lead with a BNC-BNC adaptor, or change the BNC to MCX pigtail, which is a chassis BNC socket, to one with a BNC male flying plug.

The MCX connector plugs into the RTL SDR dongles antenna socket. The dongle will be removed from its plastic case and mounted bare, with its chip being fitted with a heatsink. So much for the RF side. Now the fun starts...

The RTL dongle is power hungry, ive measured its current draw at 280mA. This is more than really sensible for the R-Pi's USB ports. So, a Y-splitter cable, with two female and one male USB connector, will be used to seperate the R-Pi's USB port and allow a higher current power feed. The RTL dongle will plug into one of the female connectors, and the male connector into the R-Pi. The remaining female connector will have a USB break out module inserted, which allows me to make direct connections. To this will be connected a short microUSB plug cable, Vbus and Gnd only, this is the power feed to the R-Pi. Two more slightly heavier gauge wires will connect Vbus and Gnd of this break-out board to the output of a Buck Converter module. There should be no need to isolate the Buck Converters supply from that of the R-Pi's USB port, but I may do so anyway, this should be just a case of lifting the appropriate polyfuse on the board, or otherwise isolating the boards 5v rail from one of the USB sockets.

So thats the RF and local system power sorted. Now for the data link and PoE.

A low cost PoE splitter will connect the incoming CAT5 cable to the R-Pi's Ethernet port. The 2.1mm DC socket of this splitter will then connect to the input of the Buck Converter. This I 'might' do with a proper 2.1mm DC jack, if I can be bothered, or I may cut the connector off and hard wire it to the module. It probably doesnt matter. Its probably sensible to add a fuse at this point though.

And that, apart from a suitable waterproof cable gland, a sturdy radome and some pole mounting hardware, is the head end system in its entirety!

All that remains then is the 'interface' end - where the system connects to the LAN and hence on to the internet. This is really very simple. The other half of the PoE pair - the Injector - will connect between the CAT5 cable run and the router. A suitably rated DC power supply then feeds into the injector, which feeds this up the unused 4,5/7,8 cores of the CAT5 cable. Here we have to account for the resistive losses of the run length, and the current maximum ratings of the cable, which is 577mA per core.

At the head end, we want to be able to provide a minimum of 10W. As this is a 5V system, that means 2A. Clearly this cant be done within the cable rating even ignoring resistance loss. This of course is why we have the Buck Converter!

As we have two cores to carry supply, we have a maximum current capacity of just over 1.1A. But i'd rather derate the cable to no more than 500mA per core. Using a 12V supply we can have 416mA per core, even better is to go up to 24V or even, if possible 48V, where we would have little over 100mA per core, for a 10W power total. So anywhere between 24-48V will be ideal, depending on what power packs I can get my paws on. Ideally, a mod to the Router will be made to allow the router and PoE all to run on just one efficient PSU.

So all the necessary parts are either in stock or on order, other than a PSU, and a suitable housing. I suspect however that will end up being PVC pipe, as I can get suitable end caps etc, and it will all seal nicely.

In the meantime, while I await delivery of the parts, I have a 1:72 scale Sopwith Camel to build!

Sorry, no pictures with this post! But hey, have £50 on me -

On each blog post  I will be including the following link -
 £50 credit if you switch energy supplier to Bulb Ltd
This is my own personal referral link, not an advert! I am with Bulb for gas and electricity. Do your own research, but if you decide to switch to them, do so via the above link, and we both get a £50 credit!


Sunday, 18 February 2018

ADS-B reception again

Its been some time since I played with receiving ADS-B signals from aircraft, but with the prospect of being able to track military flights in return for contributing data, I looked again at the 360radar website.

In order to contribute, you of course need to be able to receive, decode, and upload the ADS-B 'squitters'. The easiest way to do this for us was to make use of one of the RTL SDR dongles I have, plus Sams Raspberry Pi 2B single board computer


The 360radar website contains all the instructions needed to install the necessary software and get the system working, so between me and Sam we followed this and soon had decoded signals from a slack handful of aircraft using the dongles crappy stock magmount antenna stuck to the top of a Horlicks tin.

The next job then was to improve the antenna. I decided on  a simple 1/4 wave ground-plane design, but utilised twelve radials over the normal four, the resulting ground-plane at the frequency in question, 1.09GHz, being to RF all but solid. This antenna was built around a spare panel mount BNC socket

What was awkward was soldering the elements on! A 45W iron is not quite powerful enough for the job! It worked, but was hard to do. Ive ordered a 150W iron for any future heavy tasks like this!

With the new antenna just held close to vertical temporarily the resulting increase in received aircraft was astounding! from around 5-8 to 50-60!


However, since all this was rigged on Sams windowsill, it couldnt stay this way. So I had to create a stand to mount the antenna on, so that it can remain in use for some time, until we get the system compact and contained for mast-head mounting.


Theres nothing flash about this mount. Its literally a bit of PVC conduit, rammed into a hole in a block of wood, with a BNC to BNC patch lead in it!


And here it is with the antenna mounted and the receiver connected. It keeps it vertical, and more to the point tidy. The screenshot below shows a moment in the decoded aircraft signals using this antenna.


More work is required, for a start I need to let Sam have his R-Pi back!  So the software will need to be installed on another R-Pi that im arranging to obtain, and the whole thing - antenna, receiver and computer, mounted into a weatherproof radome for installing on the roof apex mast. I plan on running another length of external grade CAT5 cable for this, and making use of PoEt (Power over Ethernet), so I also have some PoEt adapters on order.

Ive received the amplifiers that I intend to use with the HB100 radar modules as well, so will need to fit that experiment in soon.


On each blog post  I will be including the following link -
 £50 credit if you switch energy supplier to Bulb Ltd
This is my own personal referral link, not an advert! I am with Bulb for gas and electricity. Do your own research, but if you decide to switch to them, do so via the above link, and we both get a £50 credit!

Friday, 2 February 2018

First Experiments with HB100 Doppler RADAR Modules

 Cor blimey! Aint it cold!!!

These past few days here have been enough to freeze the balls off of a brass monkey! (although monkeys were never made of brass for this exact reason - if you dont know what a 'monkey' is, research naval gunnery in the 18th century!). If your finding that your gas bill is extortionate this time of year, take a look at the link at the bottom of this page!

The HB100 Doppler RADAR module is the modern equivalent of the old X-band Gunn diode modules that many of us used for 10GHz WBFM back in the 80's and 90's. And they cost very little. I have two of these to experiment with.

The ultimate goal of course is to use them for 10GHz WBFM amateur radio, but they are somewhat harder to use in this manner than the old Gunn modules were.

Instead, as a project for my son who is well into his coding, we are looking at a simple RADAR speed gun project. To this end, I had a little play with one yesterday.

Nothing very special, and no pictures im afraid for this post. Simply a test that they do in fact work. I took one of the modules and connected its IF output to the 'scope, and gave it a clean regulated 5V supply. The output is extremely low level, we are talking just a handful of mV here, and is a variable audio frequency, dependent on the speed of the target, which in this case was just my hand being waved about! But - I could detect an output waveform.

Next step is to add a suitable IF amplifier chain to bring the output level up to around 3V or so, suitable for shaping and feeding into an Arduino.

Also yesterday, I finally joined the Milscanners group and programmed by UBC-125XT scanner up with the UKs military airband frequencies that are most likely to be receivable from the shack. Im quite surprised to be able to hear aircraft at low level all the way down to RAF Cranwell, and extremely pleased to be able to hear at good level aircraft at the Donna Nook gunnery range. The trick now will be to identify any frequencies that I just cant get any traffic on and remove them - as there are still around 60 air defence channels I want to add but have run out of memories!



On each blog post  I will be including the following link -
 £50 credit if you switch energy supplier to Bulb Ltd
This is my own personal referral link, not an advert! I am with Bulb for gas and electricity. Do your own research, but if you decide to switch to them, do so via the above link, and we both get a £50 credit!

Sunday, 28 January 2018

Converting the LED Emergency Lights

Well the Buck Converter modules arrives a couple of weeks ago, but its been too cold to be in the workshop for long, so ive only just got around to progressing this,

The original plan was to remove ALL components from the original boards, both through-hole and SMT, except for the Constant Current Driver. This proved to be impossible as the SMT parts are all glued on! So instead, I decided to isolate the driver IC by milling away the track and the SMT parts in the area around it!

This first board you can see the result of attempting to remove the SMT parts with the heat gun!
I also decided to keep one of the fuse holders in place for the 12v feed to the Buck Converter. Using the original track on the board allowed short bridge wires to be used to connect the module, with a bit of hotmelt glue to hold it in place


The 10 turn preset for setting the output voltage seems to have far too many turns than required - it took a fair bit of 'screwing down' to reach the 3.9V required. Next step is to connect this to the Driver IC, add any other required parts (bypass capacitors etc), replace any poor output LEDs, and wire it all up. An on/off switch will be added to the casing.

Although I do have a lot of 5mm white LEDs, Ive also started to salvage the ones in my now failed Black & Decker LED worklight.

As Sam is becoming a very competent programmer, we've also decided on another Arduino project. As I obtained a couple of HB100 10GHz doppler radar modules some time back, we are going to build a speed gun!



On each blog post  I will be including the following link -
 £50 credit if you switch energy supplier to Bulb Ltd
This is my own personal referral link, not an advert! I am with Bulb for gas and electricity. Do your own research, but if you decide to switch to them, do so via the above link, and we both get a £50 credit!